EVALUATING PRESTRESSING STRANDS AND POST-TENSIONING

CABLES IN CONCRETE STRUCTURES USING

NONDESTRUCTIVE METHODS

By

Dr. Elin Jensen, Principal Investigator

Dr. Nabil Grace, Co-Principal Investigator

Mr. Samuel Kumi, Research Assistant

Department of Civil and Architectural Engineering Lawrence Technological University

Southfield, MI 48075-0134

November 2015

TECHNICAL REPORT DOCUMENTATION PAGE

1. Report No. RC-1632

2. Government Accession No. N/A

3. MDOT Project Manager Beckie Curtis

4. Title and Subtitle Evaluating Prestressing Strands and Post-Tensioning Cable in Concrete Structures using Nondestructive Evaluation Methods

5. Report Date

6. Performing Organization Code N/A

7. Author(s) Dr. Elin Jensen, Dr. Nabil Grace, and Mr. Samuel Kumi

8. Performing Org. Report No. N/A

9. Performing Organization Name and Address Lawrence Technological University 21000 West Ten Mile Road Southfield, Michigan 48075

10. Work Unit No. (TRAIS) N/A

11. Contract No. 2010-0293

11(a). Authorization No. Z4

12. Sponsoring Agency Name and Address Michigan Department of Transportation Research Administration 8885 Ricks Rd. P.O. Box 30049 Lansing MI 48909

13. Type of Report & Period Covered Final Report 10/01/12 to 6/30/2015 14. Sponsoring Agency Code N/A

15. Supplementary Notes

16. Abstract The objectives were to evaluate the ability of different NDE methods to detect and quantify defects associated with corrosion of steel reinforcement and grout defects in post-tensioning applications; and to evaluate the effectiveness of selected NDE methods to detect and quantify the deterioration. The study included side-by-side box beam bridges for evaluation of the condition of prestressing strands and web reinforcement. Decommissioned box beams were tested in the laboratory to relate the quantified deterioration to the measured residual flexural capacity. Also a segmental bridge was included for evaluation of the condition of the web reinforcement and post-tensioning ducts. Laboratory specimens simulated debonding and voids around steel reinforcement, loss of cross section area of steel reinforcement, and various grouting challenges. The NDE methods used were 1) ultrasonic testing for delamination/void detection; 2) electrochemical measurements for assessing the corrosive environment; and 3) magnetic flux leakage (MFL) to estimate loss of cross sectional area of embedded steel. Ultrasonic echo-pulse tomography was able to detect debonding of the steel reinforcement and to detect internal voids and delamination. The method was able to quantify the depth of the defect but not the defect length. Electrochemical testing isolated the areas with high chance of corrosion. The MFL pilot demonstration showed that MFL can detect changes in the area of the steel reinforcement in laboratory and field applications. Presently, the calibration overestimates the steel loss. There was good agreement between estimated beam flexural capacity based on NDE results and the measured capacity of full scale box beams. 17. Key Words Corrosion, Concrete, Prestressed, Post-tensioning, Nondestructive testing, ultrasonic testing, magnetic flux leakage, laboratory and field investigation

18. Distribution Statement No restrictions. This document is available to the public through the Michigan Department of Transportation.

19. Security Classification - report Unclassified

20. Security Classification - page Unclassified

21. No. of Pages

22. Price N/A

iii

DISCLAIMER

The contents of this report reflect the views of the authors who are responsible for the facts and

accuracy of the data presented herein. The contents do not necessarily reflect the official views

or policies of the Michigan Department of Transportation, nor Lawrence Technological

University. This report does not constitute a standard, specification, or regulation. Trade or

manufacturer’s names, which may appear in this report, are cited only because they are

considered essential to the objectives of the report. The United States (U.S.) government and the

State of Michigan do not endorse products or manufacturers.

iv  

TABLE OF CONTENT

TECHNICAL REPORT i

DISCLAIMER iii

EXECUTIVE SUMMARY xxii

CHAPTER 1: INTRODUCTION 1

1.1 General 1

1.2 Statement of Problems 2

1.3 Motivation 3

1.4 Study Objectives 3

1.5 Scope of Study 4

CHAPTER 2: LITERATURE REVIEW 6

2.1 Application of Non-destructive Test Methods for Assessment

of Concrete Structures 6

2.2 Visual Methods 7

2.3 Mechanical Wave Methods 8

2.3.1 Ultrasonic Testing 10

2.3.2 Acoustic Emission 15

2.4 Magnetic Methods 16

2.4.1 Magnetic Flux Leakage (MFL) 16

2.4.2. Ground Penetrating Radar 19

2.5 Electrochemical Methods 21

2.5.1 Half-cell Potential 22

2.6 Summary 25

CHAPTER 3: EXPERIMENTAL METHODS 26

3.1 Introduction 26

3.2. Construction of Two Samples of Specimen 1 Box Beams, S1-1 and S1-2 28

3.2.1 Design and Construction of Formwork. 28

3.2.2 Reinforcement Cages 30

v  

3.2.3 Simulation of Defects 32

3.2.4 Prestressing Steel Strands 34

3.2.5 Concrete Placement and Preparation of Cylinders 35

3.2.6 Curing Box Beams and Releasing of Steel Strands 36

3.2.7 Compressive Strength of Box Beams 36

3.2.8 Ultrasonic Assessment of S1-1 and S1-2 38

3.3 Construction and Testing Phases of Three Specimen #2 Beams 40

3.3.1 Construction of Specimen #2 Beams 40

3.3.2 Ultrasonic Assessment of S2-1, S2-2 and S2-3 43

3.3.3 Calibration of Magnetic Flux Leakage (MFL) Mobile Scanner 55

3.3.4 Magnetic Flux Leakage Assessment of Specimen 2 Beams

(S2-1, S2-2 and S2-3) 59

3.4 Construction of Specimen 3 Beams for Simulation of Grouting Defects 60

3.4.1 Construction Details (S3-1, S3-2, S3-3 and S3-4) 60

3.5 Evaluation of Salvaged Box Beams (J11, H6 and A1) 72

3.5.1 Experimental Program for Residual Flexural Testing 72

3.5.2 Electro-chemical Testing of Salvaged Beams Using Canin 79

3.5.3 In-situ Hardness Testing of Salvaged Beams Using

Schmidt Hammer 81

3.5.4 Ultrasonic Assessment of Salvaged Beams J11, H6 and A1 82

3.5.5 Magnetic Flux Leakage Assessment of J11, H6 AND A1 84

3.6 NDT Field Testing 88

3.6.1 US-131 over Muskegon River, Southbound Lanes 88

3.6.2 Side-by-side Box Beam on I-96 over Canal Road, Lansing 100

CHAPTER 4: ANALYSIS AND DISCUSSION OF LABORATORY TEST RESULTS 104

4.1 Overview of Laboratory Specimen and Testing Set-ups 105

4.2 Analysis of Ultrasonic Assessment Data Using Statistical Matched Pairs 107

4.2.1 Laboratory Case study: S1-1 and S1-2 Box Beams 107

4.3 Analysis of Ultrasonic Assessment Data Using Comparison of

Reflected Intensities 114

vi  

4.3.1 Laboratory Case Study: S1-1 and S1-2 Box Beams 114

4.3.2 Laboratory Case Study: S2-1, S2-2 and S2-3 Beams 117

4.3.3 Laboratory Case Study for Grouting Defects of S3-1, S3-2, S3-3 and S3-4 120

4.4 Analysis of MFL Data 122

4.4.1 Background 122

4.4.2 Laboratory Case Study: S2-1, S2-2 AND S2-3 Beams 127

CHAPTER 5: ANALYSIS AND DISCUSSION OF FIELD TEST RESULTS 139

5.1 Salvaged Box Beams from Kent County 139

5.1.1 Ultrasonic Assessment of Salvaged Box Beams 139

5.1.2 Electrochemical Measurements on Box Beams 148

5.1.3 Impact Hammer Test on Box Beams 157

5.1.4 Magnetic Flux Leakage Assessment of Salvaged Box Beams 158

5.1.5 Analysis of Residual Flexural Testing of Salvaged Box Beams 162

5.2 I-96 Box Beams over Canal Road 165

5.2.1 Ultrasonic Assessment of I-96 Box Beams 166

5.2.2 Electrochemical Half-cell Assessment of I-96 Box Beams 168

5.3 Segmental Box Beam Bridge carrying US-131 over Muskegon River 170

5.3.1 Ultrasonic Assessment 171

5.3.2 Electrochemical Half-cell Assessment 182

CHAPTER 6: SUMMARY AND CONCLUSIONS 187

6.1 Research Objectives and Methods 187

6.2 Assessment of Defects Associated with Corrosion of Embedded Steel 192

6.2.1 Ultrasonic Testing 192

6.2.2 Half-Cell Potential Difference 194

6.2.3 Magnetic Flux Leakage Measurements 196

6.3 Assessment of Grouting Challenges in Post-tensioning Ducts 198

6.3.1 Laboratory Simulations 198

6.3.2 Field Explorations 199

vii  

APPENDIX A:

Condition Assessment of Side-by-Side Box Beam Bridge Carrying US-24 over

Middle Rouge River 201

APPENDIX B:

Implementation Plan 225

APPENDIX C:

Design Calculations for Specimen1 Laboratory Box Beams 231

APPENDIX D:

Sample Calculations for Residual Strength of MDOT Salvaged Box Beam (J11) 265

APPENDIX E:

Scan Results of Ultrasonic Assessment of Grouting Defects of S3-1, S3-2, S3-3 & S3-4 273

APPENDIX F:

Magnetic Flux Leakage Assessment Graphs for Salvaged Box Beams J11, H6 & A1 281

REFERENCE 303

viii  

LIST OF FIGURES

Fig. 2.1: Longitudinal Wave Propagation (Top) and Transversal Wave Propagation 9

Fig. 2.2: Conversion of a P wave striking an interface between dissimilar materials,

following Snell’s law 10

Figure. 2.3: Configuration of a Transmitter and Receiver. 11

Figure 2.4: Schematic of pulse velocity device (Crawford, 1997) 12

Figure 2.5: Illustration of the Basis of Ultrasonic Pulse Echo Method 13

Figure 2.6: Illustration of the Basis of the IE method. (Pessiki et al, 2010) 15

Figure 2.7: Change in magnetic flux before and after detection of flaw (defect) 18

Figure 2.8: Propagation of EM energy through dielectric boundaries 20

Figure 2.9: Setup for electrochemical techniques 22

Figure 2.10: Schematic view of the electric field and current flow of an active/passive

macro-cell on steel in concrete 23

Figure 3.1: Specimen 1 Box Beams, S1-1 and S1-2 26

Figure 3.2: Specimen 2 Beams, S2-1, S2-2 and S2-3 27

Figure 3.3: Specimen Beams, S3-1, S3-2, S3-3 and S3-4 27

Figure 3.4: Construction of formwork. 29

Figure 3.5: Construction of interior cells 29

Figure 3.6: Formwork set-up 29

Figure 3.7: Cross-sectional details of box-beams 30

Figure 3.8: Longitudinal details of box-beams 31

Figure 3.9: Bending stirrups 31

Figure 3.10: Tiring and aligning reinforcement 31

Figure 3.11: Rebar cages for the box-beam 31

Figure 3.12: Installing Edge & Side Panels 31

Figure 3.13: Plastic defects around Top Rebar 32

Figure 3.14: Plastic defects around Bottom Strand 32

ix  

Figure 3.15: Honeycomb at the top of box beam 32

Figure 3.16: Plastic Tubes for Debounding 32

Figure 3.17: Top Plan of Box Beam 1 showing locations and defect Length around

Top Rebar 33

Figure 3.18: Bottom Plan of Box Beam 1 showing locations of defects around strands 33

Figure 3.19: South Elevation of Box Beam 1 showing locations of defects around strands 33

Figure 3.20: End Plates for prestressing bed 34

Figure 3.21: Prestressing steel strands 34

Figure 3.22: Casting of Box Beams, S1-1 and S1-2 35

Figure 3.23: Slump Test / Cylinders Preparation 35

Figure 3.24: Curing Beams 36

Figure 3.25: Releasing of steel strands 36

Figure 3.26: Four (4) Days at Transfer 37

Figure 3.27: 7 Days Compressive Strength 37

Figure 3.28: Scanning on the side of S1-1 39

Figure 3.29: Scanning on top of S1-2 39

Figure 3.30: Scanning on the bottom of S1-2 39

Figure 3.31: Constructing formwork 40

Figure 3.32: Cross-sectional area loss on rebar 40

Figure 3.33: Grinding configuration of 1inches-1inches-1inches for different percentage

losses for rebar 40

Figure 3.34: Grinding configuration of 5inches-2inches-5inches for different percentage

losses for rebar 41

Figure 3.35: 0.6” Strand length of wire cut 41

Figure 3.36: Wire cut covered with dual wall Polyolefin tube and duct tape 41

Figure 3.37: Cages of rebar and strands 42

Figure 3.38: Cross-sectional Area loss 42

x  

Figure 3.39: Strands showing loss of wires covered with dual wall polyolefin tube

and duct tape 42

Figure 3.40: Placing of concrete 43

Figure 3.41: Leveling top of specimen 2 beams 43

Figure 3.42: Checking for slump 43

Figure 3.43: Preparation of concrete cylinders 43

Figure 3.44: Top Rebar Arrangement and Percentage Losses for S2-1 with

Total Length of 3 inches Cross-sectional Area Loss 45

Figure 3.45: Bottom Rebar Arrangement and Percentage Losses for S2-1 with

Total length of 3 inches Cross-sectional Area Loss 46

Figure 3.46: Cross-sectional Area of S2-1 showing Single Bottom Reinforcement 47

Figure 3.47: Top Rebar Arrangement and Percentage Losses for S2-2 with

Total Length of 12 inches Cross-sectional Area Loss 48

Figure 3.48: Bottom Rebar Arrangement and Percentage Losses for S2-2 with

Total Length of 12 inches Cross-sectional Area Loss 49

Figure 3.49: Cross-sectional Area of S2-2 showing Double Bottom Reinforcement 50

Figure 3.50: Top Strand Arrangement, Length and Number of Wire cuts for S2-3 51

Figure 3.51: Bottom Strand Arrangement, Length and Number of Wire cuts for S2-3 52

Figure 3.52: Cross-sectional Area of S2-3 showing Double Bottom Strands 53

Figure 3.53: Ultrasonic assessment set-up for S2-1 54

Figure 3.54: Scan on Rebar with defect at frequency of 70 kHz with dB of 33.8 54

Figure 3.55: Scan on rebar without defect at frequency of 70 kHz with dB of 33.8 55

Figure 3.56: MFL Mobile Scanner Calibration Setup 56

Figure 3.57: Magnetic flux leakage influence with depth 57

Figure 3.58: Graph of voltage change for each depth 58

Figure 3.59: Graph showing effect of 20% loss 58

Figure 3.60: MFL Mobile Scanner Assessing Cross-sectional Area loss or

Number and Length of Wire cuts for Specimen #2 Beams 59

xi  

Figure 3.61: Formwork Construction 60

Figure 3.62: Configuration of S3-1 showing draped post-tensioned duct with simulated

tendon paths in positive and negative moment regions and grout pumped in

to check complete grout coverage at angle breaks 61

Figure 3.63: Configuration of S3-2 showing draped post-tensioned duct to be partially

filled with water to be checked if the grout completely displaces the water. 61

Figure 3.64: Configuration of S3-3 showing straight steel duct with grout mixed with

expansion foam and plastic material 62

Figure 3.65: Configuration of S3-4 showing straight steel duct fully grouted in vertical

position to be used as a reference to assess S3-1, S3-2 and S3-3 62

Figure 3.66: Rebar preparation and installation of drapped post-tensioned duct 63

Figure 3.67: Setting Formwork 64

Figure 3.68: Beams Setup ready for casting 64

Figure 3.69: Placing Concrete into Formwork. 65

Figure 3.70: Checking Concrete Slump 65

Figure 3.71: Removal of Formwork 65

Figure 3.72: Curing of Beams 65

Figure 3.73: Grouting Equipment set-up 66

Figure 3.74: Compressive Strength Testing for Specimen #3 Concrete 68

Figure 3.75: Scanning Top of S3-1 70

Figure 3.76: Scanning Side of S3-1 70

Figure 3.77: Scanning Top of S3-2 70

Figure 3.78: Scanning Top of S3-3 70

Figure 3.79: Scanning Top of S3-4 70

Figure 3.80: Examples of Scan Reflections from Side of S3-1 at 50 kHz before and

after Grouting. 71

Figure 3.81: Examples of Scan Reflections from Side of S3-3 at 50 kHz before and

after Grouting. 71

xii  

Figure 3.82: Salvaged Box Beams from Kent County 72

Figure 3.83: Typical Cross-section of Salvaged Beams 73

Figure 3.84: Fabricating Spreader 74

Figure 3.85: Fabricating Spreader 74

Figure 3.86: Layout of Strain Gages Installed on top Surface of Beams 75

Figure 3.87: Layout of Strain Gages Installed on Fascia Side of Beams 75

Figure 3.88: Layout of Strain Gages Installed on Interior Side of Beams 76

Figure 3.89: Location of LVDT Installed on Fascia Side of Beams 76

Figure 3.90: Box Beam Residual Flexural Set-up at 4 kips Loading 77

Figure 3.91: Box Beam Residual Flexural Set-up at 8 kips Loading 77

Figure 3.92: Showing Set-up for Salvaged Box Beam J11 before and after failure 78

Figure 3.93: Showing Set-up for Salvaged Box Beam H6 before and after failure 78

Figure 3.94: Showing Set-up for Salvaged Box Beam A1 before and after failure 79

Figure 3.95: Electro-chemical Half-cell of salvaged beam 80

Figure 3.96: Schmidt Hammer Testing for MDOT Salvaged Beam #1 81

Figure 3.97: Ultrasonic Assessment of Salvaged Box Beams 83

Figure 3.98: Scanning Configuration at the Bottom of Box Beam J11 85

Figure 3.99: Scanning Configuration at the Bottom of Box Beam H6 85

Figure 3.100: Scanning Configuration at the Bottom of Box Beam A1 86

Figure 3.101: Typical Cross-section of Salvaged Beams 87

Figure 3.102: Scanning Strand #5 (S5) on North Half of Savage Beam with MFL

Mobile scanner 87

Figure 3.103: US-131 over Muskegon River (SB) 89

Figure 3.104: Plan and Elevation of US-131 over Muskegon River (SB) 89

Figure 3.105: Scanning on Transverse Post-tensioning duct within Segment 2 90

Figure 3.106: Location of Transverse Post-tensioning Duct has been highlighted 90

Figure 3.107: Typical Scan Result along Transverse Post-tensioning Duct,

showing location of duct and back reflection from thickness of Top Flange 91

xiii  

Figure 3.108: Scanning on Longitudinal Post-tensioned Duct #24 and

Anchorages at the ends 92

Figure 3.109: Cross-section of SB, showing Duct #24, DWG No. BO1-4 92

Figure 3.110: Tendon anchorage details for bottom slab, DWG No. BO1-5 93

Figure 3.111: Typical Scan along Longitudinal Post-tensioning Duct 24,

showing location of duct and back reflection from thickness

of bottom flange 93

Figure 3.112: Typical Cross-section of Segment showing Bottom Flange and

Web Rebar Details, DWG No. BO1-7 94

Figure 3.113: Segments arrangement showing location of segment 2 between

abutment A and pier 1, DWG No. BO1-1 94

Figure 3.114: MFL Testing Grid Set-up within Bottom Flange of Segment #2 96

Figure 3.115: MFL Mobile Scanner with Temporary 1/8” Wood Strips 96

Figure 3.116: Locating Longitudinal and Transverse Rebar on web of Segment #2

Using Ground Penetrating Radar (GPR) 97

Figure 3.117: Electro-chemical half-cell Assessment of selected Section of

Web along Segment #2 98

Figure 3.118: Typical instrument configuration for half-cell measurements 98

Figure 3.119: Layout of Southbound Bridge indicating Test Areas #1 and #2 99

Figure 3.120: I-96 over Canal Road, Lansing 100

Figure 3.121: Scanning along web of Box beam 3-M-11 101

Figure 3.122: Typical scan along Web of box beam with back Reflection indicating

Web thickness 101

Figure 3.123: Locating Transverse Rebar for Electro-chemical Assessment using

Rebar Detector 103

Figure 3.124: Conducting Electro-chemical Analysis on the Web of External

Beam 3-M-11 103

xiv  

Figure 4.1: Over view of Specimen 1 Testing Set-up 105

Figure 4.2: Over view of Specimen 2 Testing Set-up 106

Figure 4.3: Over view of Specimen 3 Testing Set-up 106

Figure 4.4: Typical cross-section of box beams (S1-1 and S1-2) 109

Figure 4.5: Typical scan indicating diagnostic and actual thickness of top flange of S1-1

between diaphragms with y-axis showing scan depth and x-axis indicating

scan width at 50 kHz, 40 dB 109

Figure 4.6: Typical scan on stirrup without defect of box beam S1-1 between

diaphragms with y-axis showing scan depth and x-axis indicating scan

width at 80 kHz, 35 dB. 110

Figure 4.7: Typical scan on stirrup with defect (wrapped plastic tubes around rebar)

of S1-1 with y-axis showing scan depth and x-axis indicating scan width at

80 kHz, 35 dB. 110

Figure 4.8: Rebar cage set-up showing rebar with and without defects 115

Figure 4.9(a): scan on rebar with defect 116

Figure 4.9(b): scan on rebar with defect 116

Figure 4.9(c): scan on rebar with defect 116

Figure 4.10(a): scan on rebar without defect 116

Figure 4.10(b): scan on rebar without defect 116

Figure 4.10(c): scan on rebar without defect 116

Figure 4.11: Rebar cages set-up for S2-2 & S2-3 118

Fig. 4.12(a): scan on rebar with defect, 80 kHz 119

Fig. 4.12 (b): scan on rebar with defect, 90 kHz 119

Fig. 4.13(a): scan on rebar without defect 80 kHz 119

Fig. 4.13 (b): scan on rebar without defect 90 kHz 119

Figure 4.14: Configuration of S3-1 showing draped post-tensioned duct 121

Figure 4.15: Beam S3-1, showing draped post-tensioned duct 121

xv  

Figure 4.16. Depiction of magnetic flux density lines around magnetic member

with or without cross sectional area reduction (ndt.net) 122

Figure 4.17. Variation in hall voltage output for area with or without cross sectional

area reduction. 123

Figure 4.18. Relationship between hall voltage and cross sectional area of rebar.

for concrete cover of 1.5 inches. 125

Figure 4.19. Predicted rebar cross sectional area loss versus actual rebar cross

sectional area loss over 12 inches in calibration beams. 126

Figure 4.20: Top rebar layout for S2-1 with rebar B highlighted 127

Figure 4.21: M3 Consolidated Graph for S2-1, Rebar B, Top 128

Figure 4.22: M3 Adjusted Graph for S2-1, Rebar B, Top 129

Figure 4.23: Bottom rebar layout for S2-2 with percentage cross-sectional losses 133

Figure 4.24: M3 Consolidated Graph for S2-2, Rebar H, Bottom 134

Figure 4.25: M3 Adjusted Graph for S2-2, Rebar H, Bottom 135

Figure 4.26: Top strand layout for S2-3 with number and length of wire cuts 137

Figure 5.1 Typical scans on corroded and less corroded strand locations 141

Figure 5.2 Typical scans at 100 kHz, 34.8 dB with (a) along web and

(b) at cell location. 142

Figure 5.3 Typical scans at 90 kHz, 34.8 dB with (a) along web and

(b) at cell location. 143

Figure 5.4 Typical scans at 80 kHz, 34.8 dB with (a) along web and

(b) at cell location. 144

Figure 5.5 Typical scans at 70 kHz, 34.8 dB with (a) along web and

(b) at cell location. 145

Figure 5.6: Typical scan on the side of Salvaged Box Beam J11 146

Figure 5.7 Typical scans from the side of box beam with scanned location

reference to end of beam 147

xvi  

Figure 5.8: Electro-chemical Half-cell of salvaged beam 149

Figure 5.9: Combined Electro-chemical Potential Map for Salvaged Beam 1

(J11), Test 1-3 150

Figure 5.10: Electro-chemical Potential Map for Salvaged Beam 2 (H6), Test 1 150

Figure 5.11: Relative Frequency for Salvaged Beam 2 (H6), Test 1 151

Figure 5.12: Cumulative Frequency for Salvaged Beam 2 (H6), Test 1 151

Figure 5.13: Electro-chemical Potential Map for Salvaged Beam 2 (H6), Test 2 152

Figure 5.14: Relative Frequency for Salvaged Beam 2 (H6), Test 2 152

Figure 5.15: Cumulative Frequency for Salvaged Beam 2 (H6), Test 2 153

Figure 5.16: Electro-chemical Potential Map for Salvaged Beam 3 (A1), Test 1 153

Figure 5.17: Relative Frequency for Salvaged Beam 3 (A1), Test 1 154

Figure 5.18: Cumulative Frequency for Salvaged Beam 3 (A1), Test 1 154

Figure 5.19: Electro-chemical Potential Map for Salvaged Beam 3, Test 2 155

Figure 5.20: Relative Frequency for Salvaged Beam 3, Test 2 155

Figure 5.21: Cumulative Frequency for Salvaged Beam 3, Test 2 156

Figure 5.22: Schmidt Hammer Testing for MDOT Salvaged Beam 1 157

Figure 5.23: Typical Cross-section of Salvaged Beams 159

Figure 5.24: Graph of Load against Displacement for Salvaged Box Beam 1 (J11) 163

Figure 5.25: Graph of Load against Displacement for Salvaged Beam 2 164

Figure 5.26: Graph of Load against Displacement for Salvaged Beam 3 165

Figure 5.27: Typical scan on the side of Box Beam 3-M-11 166

Figure 5.28 Typical scans from the side of box beam with scanned location

reference to East end of beam 167

Figure 5.29: Mapping out of stirrup locations using profometer from abutment

end towards the central pier locations, evidence of corrosion stains. 168

Figure 5.30: Test locations and recorded potential values along Fascia of 3-M-11 169

Figure 5.31: US-131 over Muskegon River (SB) 170

xvii  

Figure 5.32: US-131 over Muskegon River (SB) showing side elevation of

segmental box beams 170

Figure 5.33: Scanning on Transverse Post-tensioning duct within Segment #2 172

Figure 5.34: Transverse Road Surface Crack 172

Figure 5.35: Typical scans from selected transverse post-tensioning duct within

segment 2 from north abutment 173

Figure 5.36: Scanning on Longitudinal Post-tensioned Duct #24 and

Anchorages at the ends. 174

Figure 5.37: Cross-section of SB, showing Duct #24, DWG No. BO1-4 175

Figure 5.38: Tendon anchorage details for bottom slab, DWG No. BO1-5 175

Figure 5.39: Typical scans along (a) anchorage location at 400 mm (1’-4”)

from anchorage end and (22’-8”) from anchorage end, assessment

in north-bound direction. 176

Figure 5.40: Typical Cross-section of Segment showing Bottom Flange and

Web Rebar Details, DWG No. BO1-7 177

Figure 5.41: Segments arrangement showing location of segment #2 between

abutment A and pier #1, DWG No. BO1-1 177

Figure 5.42 Typical scans along rebar showing reflections from rebar 178

Figure 5.43: Typical Cross-section showing Test location along the Web of

Segmental box beam, DWG No. BO1-7 179

Figure 5.44: Visible pockets of voids on inner surface of web. 180

Figure 5.45: Typical scans along rebar showing pockets of voids around rebar

from scans taken from side web within segment 2 181

Figure 5.46: Electro-chemical half-cell Assessment of selected Section of

Web along Segment #2 182

Figure 5.47: Layout of Southbound Bridge indicating Test Areas #1 and #2 183

xviii  

LIST OF TABLES

Table 2.1 Interpretation of half-cell potential values as per ASTM C876 24

Table 2.2 Half-cell potential for different types of concrete 24

Table 3.1: Compressive Strength Development using 6 inch by 12 inch Cylinders 37

Table 3.2: Compressive Strength Development using 6 inch by 12 inch Cylinders 67

Table 3.3: Interpretation of half-cell potential values as per ASTM C876 80

Table 4.1: Checking concrete thickness of top & bottom flanges and side webs 112

Table 4.2: Localized scan on rebar and strands without defect. 113

Table 4.3: Localized scan on defects 114

Table 4.4: S2-1, Top Plan: Estimation of Cross-sectional Area Losses 132

Table 4.5: S2-2, Bottom Plan: Estimation of Cross-sectional Area Losses 136

Table 4.6: S2-3, Top Plan: Estimation of Cross-sectional Area Losses 138

Table 5.1: Interpretation of half-cell potential values as per ASTM C876 149

Table 5.2: Predicted Cross-sectional Area Loss 161

Table 5.3: Half-cell Potential difference along Fascia of 3-M-11 169

Table 5.4: Half-cell potential difference data of Test Area #1 183

Table 5.5: Average Half-cell potential difference data of Test Area #1 184

Table 5.6: Chance of rebar corrosion in Test Area 1 184

Table 5.7: Half-cell potential difference data of Test Area 2 185

Table 5.8: Average Half-cell potential difference data of Test Area 2 186

Table 5.9: Chance of rebar corrosion in Test Area 2 186

Table 6.1: Summary of Research Methods used to Detect and Quantify Defects 188

Table 6.2: Summary of Laboratory Specimens and Projects 189

xix  

Table 6.3: Description of Laboratory Specimens 190

Table 6.4: Description of Field Projects 191

xx  

xxi  

EXECUTIVE SUMMARY

Deterioration of prestressed and post-tensioned concrete members is highly related to the concrete

durability and corrosion of the steel reinforcement. In the case of prestressed structures the

corrosion causing loss of reinforcement material may also cause localized concrete damage such

as cracking, delamination and spalling. For post-tensioned structures the corrosion of the strands

is typically associated with voids in the otherwise grouted ducts. Common for both types of

reinforcement deterioration is that the onset of corrosion damage occurs without visual signs such

as cracks and stains on the concrete surface. Therefore, it is critical to be able to assess the

condition of the reinforcement to evaluate the safety of these concrete structures. Detailed

condition assessment data afford engineers the ability to make informed decisions to determine

short-term as well as long-term maintenance and repair activities. The general consensus among

researchers and practitioners is to use multiple nondestructive test methods along with visual

inspections and coring to effectively assess the structural performance.

The goal of this study was to assess the effectiveness of multiple nondestructive evaluation

(NDE) methods to evaluate the condition of prestressed concrete bridges exhibiting deterioration

associated with corrosion of the steel reinforcement. The research plan was designed around an

experimental approach and included both laboratory and field investigations. The objectives were

to (1) develop laboratory references, for selected NDE methods, for detecting and quantifying

defects associated with corrosion of steel reinforcement and grout defects in post-tensioning

applications; and (2) evaluate the effectiveness of the selected NDE methods to detect and quantify

deterioration in the field using both in service and decommissioned bridges.

Side-by-side box beam bridges were selected for evaluation of the condition of prestressing

strands and web reinforcement. Decommissioned box beams were tested in the laboratory to relate

the impact of the estimated material deterioration, by the NDE methods, and the measured residual

flexural capacity of the beams. Also a segmental bridge was included for evaluation of the

condition of the web reinforcement and post-tensioning ducts. Specialized laboratory specimens

were designed for this study to simulate debonding and voids around steel reinforcement, loss of

cross section area of steel reinforcement, and various grouting challenges. The principal NDE

methods used were ultrasonic, magnetic and electrochemical. The following primary methods

were used:

xxii  

Ultrasonic assessment for delamination, void detection and thickness measurements.

Electrochemical half-cell potential assessment for assessing the corrosive environment.

Magnetic flux leakage assessment to estimate loss of cross sectional area of steel

reinforcement.

The ultrasonic and the electrochemical assessments were conducted using commercially

available systems. The magnetic flux leakage assessment were performed using a strong earth

magnet mounted on a mobile unit. The change in magnetic field leakage as the magnet passes

over the steel reinforcement is correlated to steel loss. The pilot version of the MFL mobile system

was developed at the Center for Innovative Research at Lawrence Technological University.

Ultrasonic echo-pulse tomography was able to detect debonding of the steel reinforcement and

to detect internal voids and delamination. The method was able to quantify the depth of the defect

but not able to distinguish the debonding length around the reinforcement; unless cracking and

delamination had reached an advanced stage. Electrochemical testing isolated the areas with high

chance of corrosion. From the field evaluation of the box beams, the areas showing high chance

of corrosion coincided with areas experiencing increased ultrasonic reflections indicating various

stages of cracking and delamination around the steel reinforcement. The MFL pilot demonstration

showed that MFL can detect changes in the area of the steel reinforcement in laboratory and field

applications. Presently, the MFL calibration displays the strongest correlation between predicted

and actual loss when the loss is spread out over a longer area than when it is concentrated

(12 inches contrast to 3 inches). There was good agreement between estimated beam flexural

capacity based on NDE assessment and the measured capacity of full scale box beams.

The study proposes an implementation strategy for integrating the use of ultrasonic echo-pulse

tomography to detect and quantify defects around the reinforcement. The method is able to

evaluate the extent of deterioration along the reinforcement beyond the immediate area suffering

from extensive visual deterioration such as rust stains, spalling and cracking. Furthermore,

ultrasonic assessment can detect early deterioration in the shear key region of the box beams. The

study also proposes the use of electrochemical assessment as part of monitoring the change in the

chance of corrosion in areas that exhibit surface cracking developed during construction, early

deployment, or other events. The MFL mobile scanner used in this study is not generally field

worthy until it is running on tracks mounted on the structure or a scaffold system.

1

CHAPTER 1: INTRODUCTION

1.1 General

Detailed condition assessment data afford asset managers and structural engineers the ability to

make informed decisions to improve short-term as well as long-term maintenance and repair

activities of concrete structures. Recent advancements in nondestructive evaluation techniques

produce an excellent source of data for an informed decision.

The major cause of deterioration in prestressed strands and post-tension cables in bridges is

corrosion of prestressed strands which is usually associated with delamination and spalling as well

as grouting defects in post-tensioned ducts which also results in corrosion of post-tensioned cables.

It is against this background that a study was conducted to evaluate the condition of prestressing

strands and post-tensioning cables in concrete structures using nondestructive evaluation (NDE)

methods. The following primary nondestructive methods were deployed: ultrasonic assessment for

delamination and void detection; electro-chemical half-cell assessment for detecting corrosive

environment; and magnetic flux leakage to determine loss of cross sectional area of rebar and

strands.

Three groups of laboratory specimen with simulated defects of voids and cross-sectional area loss

were designed for testing using the above nondestructive methods. Three side-by-side field beams

from Kent County Road Commission decommissioned after 39 years of service were selected for

nondestructive evaluation and residual flexural testing. Field evaluations were also conducted on

portions of the segmental bridge carrying US-131 over Muskegon River, 6 miles south of Big

Rapids as well as segments of the side-by-side box beam bridge carrying I-96 over Canal Road in

Lansing. Results of nondestructive assessment conducted on selected box beams of northbound

US-24 over Middle Rouge River have been presented in the Appendix of this report as an

additional study.

2

1.2 Statement of Problems

Prestressed beams commonly used in highway bridge construction and segmental post-tensioned

bridges are widely used for medium and long span bridge construction. These types of structures

rely heavily on steel prestressing strands and post-tensioning cables for strength and durability.

Additionally, it is very important during the construction of these bridges to implement an effective

quality control plan to ensure proper placement of the strands and to ensure that the ducts are fully

grouted in accordance with the design. Methods to determine the overall conditions of these

strands are critical to verify the overall integrity and safety of these structures. Since the strands

are embedded in concrete, and often the area is complex and congested, nondestructive evaluation

(NDE) methods are needed to assess the condition of these structures. When the evaluation is

performed routinely the results can aid the management of the asset in terms of operation, repair

and replacement activities.

In particular, exposed structures like highway bridges are more vulnerable to environmental

deterioration. The deterioration of prestressed and post-tensioned concrete members is highly

related to the concrete durability and corrosion of the steel reinforcement. In the case of

prestressed structures the corrosion products may cause localized concrete damage leading to

delamination and spalling. In the case of post-tensioned structures the corrosion is typically

associated with regions of voids in the otherwise grouted ducts. Yet, common for both types of

structures is that the onset of damage occurs without visual clues such as cracks and stains on the

concrete surface. However, a significant indicator of the potential for ‘hidden’ deterioration is the

formation of voids. Voids may form either as a result of the corrosion product, delamination and

cracks in the prestressed concrete beam, or voids due to lack of proper grouting of the post-

tensioning ducts. A second indicator is the change in magnetic resistance of the strands due to the

loss of cross section area.

During the last decade, governments and industry have made significant investments into

developing new non-destructive testing (NDT) methods for assessing the conditions of prestressed

and post-tensioned concrete structures. The Federal Highway Administration's (FHWA) Office

of Infrastructure Research and Development has identified NDE as a primary activity in achieving

3

long term infrastructure performance as well as durable infrastructure systems (FHWA-HRT-08-

069).

At the present time, the general consensus among researchers and practitioners is to use multiple

NDT methods along with visual inspections and coring to effectively assess the structural

performance. There are different nondestructive evaluation (NDE) methods for monitoring and

instant assessing prestressed and post-tensioned concrete members. The methods range from semi-

permanently installed sensors such as Moisture Sensors, Acoustic Emission Sensors (AES) and

Micro Electro-Mechanical Sensors (MEMS) to manually operated mobile sensor systems such as

Magnetic Flux Leakage (MFL) and Ultrasonic Pulse-Echo Method.

1.3 Motivation

The research team at Lawrence Technological University (LTU) proposed a comprehensive study

to investigate and recommend NDE methods for assessing the conditions of prestressed and post-

tensioned concrete members. The study included a review of existing use of NDE methods to

evaluate the condition of prestressed and post-tensioned strands in concrete structures as well as

to verify the proper grouting of post-tensioned ducts in concrete structures. Experimental

laboratory and field investigations were performed to evaluate the NDE methods ability to detect

and quantify defects. MDOT engineers assisted in defining the selection criteria and the selection

of the candidate bridges.

The outcome of these investigations was to provide recommendations on suitable (rapid and

flexible) NDE methods to implement as an inspection technique in the bridge inspection protocol

for prestressed and post-tensioned concrete bridge systems.

1.4 Study Objectives

The main objectives of this project are as follows:

Review existing use of NDE methods to evaluate the condition of prestressed and post-

tensioned strands in concrete structures, and to verify the proper grouting of post-tensioned

ducts in concrete structures.

4

Evaluate the effectiveness of NDE methods to evaluate the condition of prestressed and

post-tensioned strands in concrete structures and verify proper grouting of post-tensioned ducts

in concrete. Detecting and quantifying defects are the primary measures used to evaluate the

methods.

Development of magnetic flux leakage device using the active MFL method.

1.5 Scope of Study

A detailed experimental investigation was designed to address the study objectives. The study

utilized laboratory prepared and field specimens tested in the laboratory as well as field inspection

on existing structures. Three sets of laboratory specimen and three salvaged box beams received

from Kent County Road Commission were evaluated in the laboratory. Two laboratory box beams

were constructed according to the geometry stipulated in MDOT Bridge Design Guides 6.65.10A

with pre-induced defects to simulate voids created around prestressed strands and post tensioned

cables as a result of corrosion by using plastic tubes as detailed in the construction phase. Three

laboratory beams were constructed with pre-induced grinding defects covered with plastic tubes

to simulate cross-sectional area loss for detection and quantification using ultrasonic assessment

and magnetic flux leakage system. Finally, four beams were constructed to simulate grouting

defects normally associated with post-tensioned cables.

Field evaluations were also carried out on sections of the following existing bridges in Michigan,

US-131 over Muskegon River, southbound 6 miles south of Big Rapids, MI, and segments of S08-

55602 of I-96 over Canal Road in Lansing, MI. The primary employed nondestructive evaluation

methods were ultrasonic, magnetic flux leakage and electro-chemical half-cell assessments. This

report consists of the following chapters:

Chapter (2): This chapter summaries existing use of NDE methods to evaluate the condition of

prestressed and post-tensioned strands in concrete structures as well as to verify

proper grouting of post-tensioned ducts in segmental concrete structures.

Chapter (3): This chapter presents details of construction of laboratory specimen as well as

methodology and testing on existing bridges in this study: US-131 over Muskegon

5

River, southbound 6 miles south of Big Rapids and segments of I-96 over Canal Road

in Lansing.

Chapter (4): This chapter presents the detailed analysis of the test results from the laboratory

investigations.

Chapter (5): This chapter presents the detailed analysis of the test results from the field

investigations.

Chapter (6): This chapter presents the summary and conclusions derived from this investigation.

Chapter (7): This chapter presents the recommendations derived from this investigation.

6  

CHAPTER 2: LITERATURE REVIEW

2.1 Application of Non-destructive Test Methods for Assessment of Concrete Structures

Many structures are built of prestressed concrete in which prestressing steel wires are put into a

permanent state of tension to compensate for the low tensile strength of the concrete. Hence, tensile

cracking in the concrete is minimized by ensuring that the concrete is in compression under normal

working loads. However, some existing prestressed concrete bridge beams experience

deterioration that raises serious concerns about the long-term durability of structures. The major

cause of deterioration in prestressed strands in bridges is corrosion. In prestressed concrete

structures, the high stress level in the tendons strongly modifies the steel corrosion process. Stress

corrosion is characterized by the coupling between the conventional corrosion (pitting attacks in

chloride environment) and the steel micro cracking; the latter induced by the high stress level and

hydrogen embrittlement (Nürnberger, 2002). Steel micro cracking can lead to the brittle failure of

the prestressing steel even at very low corrosion level and under normal service loads (Spaehn,

1977). In such case, there may be no visible warning before tendon failure.

Deterioration of post-tensioned cables in concrete bridges is also a serious concern. The leading

cause of corrosion of post-tensioned cables is insufficient grouting of post-tensioned ducts

(Mutsuyoshi, 2001). Corrosion is a significant problem in numerous structures and the associated

cost is estimated at billions of dollars every year (Singh, 2000).

Unfortunately, corrosion of prestressing steel in both prestressed and post-tensioned concrete

members is not visible by indicators such as corrosion stain and concrete cracking during the early

stages. When visual signs of corrosion of prestressing steel occur on these members, the corrosion

activities would have been in progress for some time and extensive damage could have resulted

(Grace and Jensen, 2010).

Nondestructive evaluation techniques are commonly used in the civil engineering field. The recent

increased accuracy in detecting a specific type of parameter, such as material quality and defects,

is possible due in part to the development of new testing methods, improved data collection and

data analysis, as well as the development of user friendly systems. Contrary to destructive

7  

methods, nondestructive methods provide information about material properties and structural

conditions without compromising the structure and its serviceability. At the same time, it is

paramount to select nondestructive test methods that are field worthy and effective, as well as

implement routine evaluation schedules of these structures for improved asset management

(Ghorbanpoor, 1999).

NDE technologies used for condition assessment of prestressed and post-tensioning systems can

be grouped according to their underlying methods (Azizinamini, 2012). These groups are:

Visual methods

Magnetic methods

Mechanical wave/vibration methods

Electrochemical methods

The following sections briefly presents the common methods used for condition assessment of

concrete structures with specific focus on detection and quantification of defects associated with

corrosion of reinforcement and concrete deterioration.

2.2 Visual Methods

Visual evaluation is an important assessment in non-destructive testing and it often provides

valuable information. Visual characteristics may be due to poor workmanship as well as structural

serviceability and material degradation. It aids the engineer to differentiate between the various

signs of distress types. Defects that are usually uncovered by visual evaluation are cracks, pop-

outs, spalling, delamination, color stains, surface deterioration and lack of uniformity.

Considerable details can be collected from visual evaluation to ascertain the suitability of a

structures for its intended purpose and identifying the need for follow up in-depth inspection. An

engineer carrying out a visual evaluation of a structure should have a good knowledge of applicable

structural details, technical specification, past reports of evaluations done, construction records,

details of materials used, construction methods as well as dates of construction.

Furthermore, various technologies have extended the capability of visual inspection by using

devices like fiberscopes, bore scopes, magnifying glasses and mirrors, portable and permanent

8  

video monitoring equipment, and robotic crawlers. Use of these technologies enable visual

inspection where direct perspective is impractical; for example, the bore scope provides enhanced

ability to inspect the interior portions of post-tensioned ducts for grout voiding and strand

corrosion (Azizinamini, 2012).

2.3 Mechanical Wave Methods

The most commonly used mechanical wave techniques are ultrasonic pulse velocity methods,

impact methods, and acoustic emission. These methods make use of (small amplitude) mechanical

motion that is imposed in a material. However, they differ in the way that the stress waves are

generated and on the signal processing techniques that are used (Azizinamini, 2012).

The two main modes of wave propagation in a solid are longitudinal (compression) waves and

transverse (shear) waves. In addition, the mechanical surface excitation will generate other tertiary

surface waves. The mechanical wave methods discussed in this section are based on the behavior

of the compression and the shear waves. The direction of the particle movement in these two

independent modes are shown in Figure 2.1. The compression waves moves faster (higher

velocity) than the shear waves and therefore, these wave forms are often referred to as the primary

and secondary waves, respectively. The typical primary wave velocity for good quality concrete

is about 4000 m/s (13,100 ft./sec) and the typically secondary wave velocity is about 2500 m/s

(8,200 ft./sec).

9  

Fig. 2.1: Longitudinal Wave Propagation (Top) and Transversal Wave Propagation (Bottom)

(From Ultrasonic Technique for the Dynamic Mechanical Analysis of Polymers, BAM, 2007)

In a homogeneous, isotropic elastic material, the compression wave velocity, Vp, and the shear

wave velocity, Vs, can be expressed in terms of the elastic modulus, E, density, ρ and Poisson’s

ratio, .

and

The mechanical wave technique for thickness and void detection is based on the fundamental

concept known as Snell’s Law. Assume that a compression (P) wave is imposed on a material

with velocity V1 and the P wave strike an interface with another material with different wave

velocity V2 (e.g. concrete and air). The incident P wave will be reflected and transmitted at the

interface as P and S waves, respectively (See Figure 2.2). Snell’s law governs the relationships

between the angles of the incident, reflected and transmitted waves.

(2-1)

(2-2) 

10  

Fig. 2.2: Conversion of a P wave striking an interface between dissimilar materials,

following Snell’s law (Mehta and Monteiro, 2014)

In addition to the reflection and transmission of waves, the waves in real materials are absorbed

and scattered. This phenomena is denoted attenuation. The scattering of waves result from the fact

that the propagating medium (like concrete) is not homogenous at the micro level. It contains

internal boundaries at which the waves changes abruptly because paste and aggregates have

different velocities (elastic properties) (KrautKramer, 1990). Scattering is the reflection of the

wave in directions other than its original direction of propagation. Absorption is the conversion

of the wave energy to other forms of energy. Therefore, in real materials, the amplitude of the

reflected wave is lower than the amplitude of the incident wave.

Generally, a reliable value of attenuation can only be obtained by determining the attenuation

experimentally for the particular material being used (NDT Resource Center, 2013). This can be

done by varying the frequency during testing, especially at locations where the internal component

of the structure is known to compare how best the internal structure matches with the scanned data.

2.3.1 Ultrasonic Testing

Ultrasonic inspections in concrete are used to detect flaws (such as voids) as well as to determine

dimensions and material properties (Green, 1987; Bray and McBride, 1992). Ultrasonic inspection

11  

consists of sending and receiving waves in materials and measuring the associated travel time of

the waves. The wave frequency typically used for concrete structures is in the range of 20 to 150

kHz, however 50 to 60 kHz is suitable for most concrete applications. Low frequencies can be

used for very long concrete path lengths and high frequencies can be used for mortars or for short

path lengths. High frequency pulses have a well-defined onset but, as they pass through the

concrete, become attenuated more rapidly than pulses of lower frequency (International Atomic

Energy Agency, Vienna, 2002).

Condition and properties of the test material are determined by analyzing various properties of the

sent and received waves. Typical schematics of ultrasonic inspection systems are shown in Figure

2.3 and the configurations are classified as direct and indirect transmission between the transmitter

and receiver.

Fig. 2.3: Configuration of a Transmitter and Receiver for (a) Direct and (b) Indirect

Transmission (Mehta and Monteiro, 2014)

Several concrete variables can affect the wave velocity in concrete such as level of hydration

(porosity), moisture conditions, amount and type of aggregates, micro-cracking and presence of

reinforcement (Mehta and Monteiro, 2014). It is important to be aware of the influence of these

variables on the results.

Examples of nondestructive testing methods based on the ultrasonic testing principle are pulse

velocity, ultrasound pulse echo, impact echo, and acoustic emission. They are based on the

12  

principle of sending and receiving waves in a material where condition and properties of the test

material are determined by analyzing various properties of the sent and received waves. The

methods differ in the way that the stress waves are generated and on the signal processing

techniques that are used.

2.3.1.1 Pulse Velocity Test

Pulse velocity testing is based on measuring the travel time of the compression wave between the

transmitter and receiver. The direct transmission method is preferred in pulse velocity testing with

sensors placed on either side of the concrete and immediately opposite each other (see Figure 2.3).

The pulse generated by the transmitter causes waves of several modes but only the travel time of

the fastest traveling wave form, the compression wave, is measured. The contact between the

concrete and transmitter/receiver is ensured using a liquid coupling material such as cellulose

paste. (International Atomic Energy Agency, 2002). A schematic of the pulse velocity system is

shown in Figure 2.4.

Fig 2.4: Schematic of pulse velocity device (Crawford, 1997)

The pulse velocity, V, is given as the simply ratio of path length, L, of the wave divided by the

time, T, it takes for the wave to traverse that path length as shown in equation 2-3

(2-3)

13  

Pulse velocity testing is primarily used to determine:

the uniformity of the concrete expected to have the same properties

any concrete property changes over time

the concrete quality by correlating the pulse velocity and the concrete strength

the elastic properties (elastic modulus and Poisson’s ratio)

It is challenging to determine defects using the pulse velocity testing. When compression wave

strikes the interface between concrete and air, defects larger than the width of the

transmitter/receiver as well as longer than the wavelength of the compression wave may be

detected. In general, defects larger than about 4 inches (100 mm) can be detected (International

Atomic Energy Agency, 2002).

2.3.1.2 Ultrasound Pulse Echo

Ultrasound pulse echo testing can be used to detect objects, interfaces, and anomalies. A major

advantage is the application of the indirect setup requiring only a one sided access to the concrete

surface. If the wave velocity, V, is known, the wave travel time, t, between the transmitter, defect

and receiver is used to directly estimate the depth of the defect, d. The depth of the defect from

the surface of the concrete as shown in equation 2-4:

(2-4)

The illustration in Figure 2.5 shows the transmitted and reflected wave path. This reading is

denoted as A-scan. The method is also denoted pitch-catch method.

Fig 2.5: Illustration of the Basis of Ultrasonic Pulse Echo Method

(From NDT Systems catalog, Germann Instrument, 2010)

14  

Ultrasound Pulse Echo testing is often conducted at an ultrasound frequency of 50 kHz because of

the scattering of the sound waves by the aggregates and air pores (Nead et al, 2013). Care should

be taken during testing and analysis with the adjustment of the measuring frequency to reduce the

amount of attenuation.

It is recommended to use an array of a minimum of 10 transducers (transmitter/receivers) when

using this method for thickness assessment. A mathematical method is used to determine the true

depth by use of a phase shift superposition technique to determine the maximum synthetic echo.

The ultrasound pulse echo is capable of assessing defects in concrete elements, debonding of

reinforcement bars, shallow cracking, and delamination. The method was also successfully used

in the detection of material interfaces, based on phase evaluations of the response. Examples

include the interfaces between concrete and steel (e.g., reinforcement) or concrete and air (e.g.,

grouting defects) (Taffe and Wiggenhauser 2006; Afshari et al. 1996; Krause et al. 2008; Hevin et

al. 1998). Very shallow flaws may remain undetected because the surface waves mask the needed

compressional or shear-wave signals. Also, as ultrasonic pulse echo works with lower frequencies,

some of the defects might remain undetected (Nead et al, 2013).

Investigating ducts is sometimes complicated, in particular for structures with thick concrete cover

greater than 4 inches (100 mm). Comprehensive analysis of many A-scan readings, using 3D

SAFT (Synthetic Aperture Focusing Technique), produces a 3D topography image and the position

of the duct and potential voids can be identified.

2.3.1.3 Impact-Echo

The impact echo (IE) method is a seismic or stress wave–based method used in the detection of

defects in concrete, primarily delamination (Sansalone and Carino 1989). The objective of the

method is to detect and characterize wave reflectors or “resonators” in a concrete. This is achieved

by striking the surface of the tested object and measuring the response at a nearby location (Nead

et al, 2013).

15  

The concrete surface is struck (e.g. steel balls), and the response is measured by a nearby contact

or air-coupled sensor. The position of the defects or reflectors is obtained from the frequency

spectrum of concrete’s response to an impact (Nead et al, 2013).

This is an effective method of locating defects in piles, caissons, and plates. A mechanical impact

produces stress waves of 1 to 60 kHz. The wavelengths from 50 mm to 2000 mm propagate as if

in a homogeneous elastic medium. The mechanical impact on the surface generates compression,

shear and surface waves. Internal interfaces or external boundaries reflect the compression and

shear waves. When the waves return to the surface where the impact was generated, they can be

used to generate displacements in a transducer and subsequently, a display on a digital

oscilloscope.

The transducer acoustically mounted on the surface receives the reflections created from the

energy impact. The amplitude spectrum is then analyzed to determine if any discontinuities are

present (Pessiki et al, 2010). Figure 2.6 illustrates the basis of the IE method.

Figure 2.6: Illustration of the Basis of the IE method. (Pessiki et al, 2010)

2.3.2 Acoustic Emission

Acoustic emission is based on analyzing the response of the material during crack propagation.

The above discussed stress wave methods are different in that they are based on analyzing the

response of the material when subjected to a known signal.

16  

Acoustic Emission (AE) is defined as the elastic energy released from materials which are

undergoing deformation and crack propagation. The rapid release of elastic energy, the AE event,

propagates through the structure to arrive at the structure surface where a transducer is mounted.

These transducers detect the displacement of the surface at different locations and convert it into

a usable electric signal. By analysis of the resultant waveform in terms of feature data such as

amplitude, energy and time of arrival, the severity and location of the AE source can be assessed

(Hellier, 2001).

The main problem in field application is to distinguish the emission caused by defects from

ambient emissions. Only breaks occurring after the installation of equipment can be detected. For

passive monitoring, vibration sensors are strategically mounted on structural members. These

vibration sensors are monitored to “listen” for fracture events due to corrosion of wire strands in

the tendons (Azizinamini et al, 2012). Furthermore, severe attenuation of elastic waves in strands

bonded with concrete substantially reduces the method's effectiveness (Ciolko et al, 1999).

2.4 Magnetic Methods

Magnetic methods make use of the interaction between magnetic (and associated electric) fields and

their interaction with matter. The magnetic based cover meters have been widely used in the concrete

industry for locating the steel reinforcement and to determine the concrete cover over the

reinforcement. In the last decades the application of the magnetic flux leakage (MFL) method has

been further developed to include the estimation of loss of steel reinforcing cross section area due to

corrosion (Azizinamini, 2012). While this is a very promising technique that would allow the engineer

to quantify reinforcement loss, standardized equipment is not readily available.

2.4.1 Magnetic Flux Leakage (MFL)

The use of MFL for inspecting steel in prestressed concrete members was studied by Kusenberger and

Barton (1981) and Sawade and Krause (2007) and was later extended to other on-site inspections

(Grosse 2007). Ghorbanpoor et al (2000) developed an MFL sensor using permanent magnets to

magnetize steel components in concrete and used the amplitude of detected MFL signals to determine

empirically the flaw volume. DaSilva et al. (2009) also reported the use of MFL to estimate the amount

of material loss of corroded steel strands in concrete. Studies performed have also shown that the

17  

methodology for detecting corrosion in post-tensioned concrete bridges is promising

(Ghorbanpoor et al, 2000).

Instrumentation development has been subjected to evaluation and upgrade over the years. Various

data analysis techniques have been developed to aid the interpretation of relevant test results. Flaws

were recognizable when the flaw size was larger than approximately 10 percent of the cross-

sectional area of the specimen. With flaws smaller than 10 percent of the cross sectional area, the

correlation method, a signal analysis method based on the correlation concept was shown to be

effective (Ghorbanpoor et al, 2000). A study comparing residual magnetic field measurements to

magnetic flux leakage measurements, as a method to detect broken prestressing steel, was

conducted by Makar et al (2001). According to this investigation, analyses of two and three

dimensional magnetic field plots show that the residual magnetic field technique is capable of

detecting corrosion of a single wire in a seven-wire strand embedded in concrete with a maximum

concrete cover of 2.5 inches. MFL method has been used in Germany and reported that the

parameters associated with fractured wires are quantitatively identifiable in the laboratory (Scheel

et al, 2003). They also reported application of the method in the field.

Magnetic flux leakage for assessing condition of steel in concrete structures utilizes the

ferromagnetic property of the steel to detect disturbance of an externally applied magnetic field

due to the presence of flaws in the steel. If the direction of the applied magnetic field is set to be

collinear with the longitudinal prestressing steel, the flux lines will also be collinear with the steel

elements in the concrete. Any change in cross sectional area of the steel at any point will cause a

flux leakage, where the continuous flux lines within the steel will be forced into the medium

surrounding the flaw as shown in Figure 2.7. Since the effect of concrete on magnetic field is

negligible, the field leakage may be measured by Hall sensors in the air near the surface of the

concrete. If the magnetic source and sensors are moved along the length of the concrete member,

the changes in the field, due to the presence of flaws in steel, can be recorded as continuous in

terms of time and the field amplitude. This is then analyzed to obtain information relevant to the

location and extent of the flaw in the steel (Ghorbanpoor, 1999).

18  

Fig 2.7: Change in magnetic flux before and after detection of flaw (ndt.net)

Magnetism is caused by the presence of molecular current loops, which is caused by two

phenomena: the motion of electrons within the atoms and the spin direction of the electrons within

the magnet material. In the presence of a magnetic field, a moving charge has a magnetic force

exerted upon it. The relationship that governs the force on moving charges through a magnetic

field is called Faraday’s Law and implies that as a charge q moves with a velocity V in a magnetic

field B, it experiences a magnetic force F.

(2-5)

where:

F is magnetic force (newton)

q is charge (coulomb)

V is velocity (meter/second)

B is applied magnetic field (weber/sq. meter)

When a magnetic field comes near ferromagnetic (steel) material, the magnetic flux lines will pass

through the steel bar because the steel offers a path of least resistance due to high magnetic

permeability compared to the surrounding air or concrete. When discontinuities of defects

(corrosion) are present, this low resistance path becomes blocked and the remaining steel may

become saturated, forcing some of the flux to flow through the air. Total saturation is not required

for detection as orientation of the magnetic field can be altered by even small levels of corrosion.

However, there is a saturation level at which all dipoles are aligned and no further alignment is

possible. At saturation level, the following relationship is valid:

19  

(2-6)

where:

B is Magnetic flux (weber/sq. meter)

µ is Magnetic permeability of the material (weber/ampere meter)

H is Magnetic field strength (ampere/meter)

To use the concept of MFL as a non-destructive evaluation (NDE) tool, the device must have the

ability to measure changes in the path of magnetic field force lines near a ferromagnetic material.

Such changes in the components of the flux can be detected by one or more sensors and can be

analyzed to determine the extent or severity of the flaw. Hall-effect sensors are often used to detect

and measure MFL. The sensors are made with semi-conductor crystals that when excited by a

passage of current perpendicular to the face of crystal, react by developing a voltage difference

across the two parallel faces. The possibility of fabrication of those sensors at any size allows the

detection of small flaws on the steel (DaSilva et al, 2009)

There are two primary methods for detecting these field anomalies: active and residual. In the

active method, the sensors are placed between the poles of the magnet and readings are obtained

as the device is passed over the specimen. In the residual system, the specimen is first magnetized;

it is important that the strands become magnetically saturated. Then the device is passed over to

read the residual magnetic field. Active MFL is appropriate when large areas of corroded regions

exist inside the ducts. However, when the corroded area is small, active MFL is no longer effective

(Marcel et al, 2009). An illustration of magnetic flux leakage method is shown in Figure 2.7.

2.4.2. Ground Penetrating Radar

Ground Penetrating Radar (GPR) is a pulse method for locating structural defects (voids and

delamination), location of steel reinforcement, pavement layer thickness and structural changes

(Loizes et al, 2006; Mehta and Moneteiro, 2014). GPR system is based on the concept of sending

an electromagnetic pulse through an antenna to the concrete surface of the testing specimen and

then recording the reflected pulses from the internal interfaces, where there is a variation in the

dielectric properties (AL-Qadi et al, 2005). It is this reflection phenomenon that allows radar

20  

images to be created. A wide range of signal frequencies are used for various practical applications

ranging between 50 MHz and 2.5 GHz. Generally, for inspection of concrete structures, antennas

with frequencies of 1 GHz are used (Hugenschmidt, 2002; Mehta and Monteiro, 2014).

2.4.2.1 Physical Principle

The two most important factors affecting the propagation of radar pulses in any material are the

electrical conductivity and the dielectric constant (Sarrakento, 1997). The behavior of a beam of

electromagnetic (EM) energy as it hits an interface between two materials of dissimilar dielectric

constants, has been shown in Figure 2.8. Part of the energy would be reflected and the remaining

goes through the interface into the second material. The intensity of the reflected energy, AR, is

related to the intensity of the incident energy, AI, by the following relationship:

Fig 2.8: Propagation of EM energy through dielectric boundaries.

(International Atomic, Agency, 2002)

, (2-7)

where

, : the reflection coefficient at the interface

/ : the wave impedance of materials 1 and 2, respectively, in ohms

or written in terms of dielectric constants, and , of material 1 and 2 respectively:

21  

,√ √

√ √ (2-8)

This gives an indication that when a beam of electromagnetic energy hits the interface between

two media of different dielectric constants, the amount of reflection , is determined by the

values of their relative dielectric constants. The dielectric constant is strongly affected by the water

content and also several other factors such as mix proportions, porosity, pore fluid solution and

shapes of particles and pores.

The result of a GPR relief is a radar gram. The radar technique allows to establish the number,

position and diameter of all steel bars existing in a surveyed structure and it also allows to locate

possible cavities in both reinforced concrete structures and masonry structures. The

electromagnetic impulse technique offers the advantage of portable equipment and the capability

to detect rapidly large areas from one surface in a completely non-destructive and non-invasive

way. One of the main difficulties with the use of GPR is that the results are very difficult to

interpret and a skilled technician must be required to produce a reliable end result (Bungey et al,

2003).

2.4.2.2 Application

Locating of reinforcing bars in concrete is one of the most widespread applications of GPR in civil

engineering (Ulriksen, 1982). Recently, GPR is also used to locate ducts in post-tensioned bridges.

However steel ducts completely reflect the radar signal since they are conductive, so that the

tendon breaks or grout defects inside the duct cannot be detected (Pollock et al, 2008). GPR has

the potential to monitor grout conditions within non-conducting (plastic) ducts. In particular, GPR

is likely sensitive to the occurrence of soft, non-setting and chalky grout or cases of water intrusion

due to the changes in dielectric constants and hence reflection (Azizinamini et al, 2012).

2.5 Electrochemical Methods

These methods are used to monitor active corrosion in concrete structures by making use of the

electrochemical basis of the process. These approaches offer potential to measure meaningful data

related to active corrosion of strand. Most electrochemical techniques use the same measurement

22  

set up. It consists of a reference electrode, a working electrode, a counter electrode, and a volt

meter. A closed electrical circuit is always required and direct electrical connection to the steel

reinforcement must be established as shown in Figure 2.9 (Azizinamini, 2012).

Fig 2.9: Setup for electrochemical techniques (Azizinamini and Gull 2012)

Two non-destructive techniques are commercially available based on the electrochemical process.

These are half-cell potential and polarization resistance. The former gives information on the 

probability of corrosion and the latter is related to the  corrosion rate. The half-cell potential

measurement is practically and widely employed to identify the presence of corrosion. 

2.5.1 Half-cell Potential

The half-cell potential (HCP) measurement is a well-established and widely used electrochemical

technique to evaluate active corrosion in reinforced steel and prestressed concrete structures. The

method can be used at any time during the life of a concrete structure and in any kind of climate,

provided the temperature is higher than 2°C (Elsener, 2003). Half-cell measurements should be

taken on a free concrete surface, because the presence of isolating layers (asphalt, coating, and

paint) may make measurements erroneous or impossible. Generally, the potential difference

between the reinforcement and a standard portable half-cell, using a Cu/CuSO4 standard reference

electrode, is measured when the electrode is placed on the surface of a reinforced concrete element.

When the reference electrode is shifted along a line or grid on the surface of a member, the spatial

distribution of corrosion potential can be mapped (Baumann 2008; Gu and Beaudoin 1998).

23  

Using empirical comparisons, the measurement results can be linked to the probability of active

corrosion. Furthermore, it is meaningful to assess the change in potential over a larger surface area

to assess if there is an increased chance for an active corrosion cell.

 

2.5.1.1 Physical Principle

The procedure for measuring half-cell potentials is comparatively straight-forward. A good

electrical connection is made to the reinforcement, an external reference electrode is placed on the

concrete surface and potential readings are taken on a regular grid on the free concrete surface.

Different microprocessor-controlled single or multiple electrode devices have been developed and

are commercially available. The best way of representing the data has found to be a color map of

the potential field and contour plot, where every individual potential reading can be identified as

a small square. Alternatively, frequency plot can be used as well.

Electro-chemical half-cell potential mapping is used to identify localized corrosion of

reinforcement which greatly improve the quality of condition assessment of concrete structures

(ASTM C 876, Elsener 1992 & 2001, and Cairns and Melville, 2003, Pradhan and Bhattacherjee,

2009). Electrons flow through the steel to the cathode, where they form hydroxide (OH-) with the

available water and oxygen. This creates a potential difference that can be measured by the half-

cell method. A schematic view of the electric field and the current flow on steel in concrete is

presented in Fig 2.10. Table 2.1 shows interpretation of half-cell potential values as per ASTM

C876.

Fig 2.10: Schematic view of the electric field and current flow of an active/passive macro-cell on steel in concrete (Elsener 2001)

24  

It is often necessary to use other data such as chloride contents, depth of carbonation, delamination

survey findings, rate of corrosion results and environmental exposure conditions, in addition to

corrosion potential measurement to formulate conclusions concerning corrosion activity of

embedded steel and its probable effect on service life of a structure. (ASTM Designation:

C876-09, Standard test method for corrosion potential of uncoated reinforcing steel in concrete)

Table 2.1 Interpretation of half-cell potential values as per ASTM C876

Half-cell potential (mV) relative to Cu/CuSO4

Chance of rebar being corroded

< -500 Visible evidence of corrosion

-350 to -500 95%

-200 to -350 50%

> -200 5%

Typical order of magnitude for the half-cell potential of steel in concrete measured against a

Cu/CuSO4 reference electrode (RILEM TC 154-EMC) is presented in Table 2.2. It is to be noted

that Table 2.1 should not be used in isolation, but should be used considering the environmental

conditions of the concrete being tested, as this has an impact on the half-cell potential value as

shown in Table 2.2.

Table 2.2 Half-cell potential for different types of concrete

Concrete condition Half-cell potential

Water saturated concrete without O2 -1000 to -900 mV

Moist, chloride contaminated concrete -600 to -400 mV

Moist, chloride free concrete -200 to +100 mV

Moist, carbonated concrete -400 to +100 mV

Dry, carbonated concrete 0 to +200 mV

Dry, noncarbonated concrete 0 to +200 mV

25  

The main disadvantage of this method is that the potentials measured are too sensitive to moisture

content, thickness of concrete cover, surface coating, resistivity of concrete, and the type of

electrode (Misra and Uomoto, 1990).

Theoretical considerations and practical experience on a large number of structures have

shown (Elsener et al, 1990, 1992 & 1979) that the results of potential mapping on existing

structures need careful interpretation. No absolute values of potential can be applied to indicate

corrosion hazard in a structure. Depending on concrete moisture content, chloride content,

temperature, carbonation of the concrete and cover thickness, surface coating, resistivity of

concrete, type of electrode, different potential values indicate corrosion of the rebar on different

structures. After repair work, interpretation of half-cell potentials can be even more difficult

because alkalinity, composition and resistivity of the pore solution in the repaired areas can be

greatly altered compared to the old existing concrete.

2.6 Summary

This chapter highlighted some of the most commonly used nondestructive testing method used for

condition assessment of steel reinforced concrete structures. The discussion was limited to topics

of visual inspection, mechanical wave methods, magnetic methods, and electrochemical methods.

The following basic principles governing the methods were briefly discussed as well as associated

applications.

Visual inspection is for surface evaluation of delamination, cracking, and stains.

Mechanical wave method is for detecting voids, defects as well as material interfaces.

Magnetic methods are for locating steel reinforcement and estimating the cross-sectional loss

of steel reinforcing material

Electrochemical methods are for assessing the chance for corrosive environment in

reinforced concrete.

26  

CHAPTER 3: EXPERIMENTAL INVESTIGATION 

3.1 Introduction

The experimental investigations were undertaken to evaluate the effectiveness of nondestructive

evaluation (NDE) methods to assess the condition of prestressed and post-tensioned strands in

concrete structures and verify proper grouting of post-tensioned ducts in segmental concrete

structures. The study utilized laboratory prepared and field specimens tested in the laboratory as

well as field inspection on existing structures. Three sets of laboratory specimen and three salvaged

beams received from Kent County Road Commission were evaluated in the laboratory.

Specimen 1 was made up of two (2) box beams constructed according to the geometry stiputed in

MDOT Bridge Guide section 6.65.10A. The box beams were simulated with voids created around

prestressing strands and transverse reinforcement using plastic tubes as detailed in section 3.2.3 of

this report. Specimen 2 was made up of two (2) beams constructed with reinforcement simulated

with grinding defects to depict cross- sectional area loss as detailed in section 3.3.1, for detection

and quantification using ultrasonic and magnetic flux leakage assessments. Specimen 3 was made

up of four (4) beams constructed with post-tensioned ducts simulated with grouting defects as

detailed in section 3.4.1 to be assessed using ultrasonic method. Laboratory prepared specimen

are shown in Figures 3.1 through 3.3.

Figure 3.1: Specimen 1 Box Beams, S1-1 and S1-2.

27  

Figure 3.2: Specimen 2 Beams, S2-1, S2-2 and S2-3

Figure 3.3: Specimen Beams, S3-1, S3-2, S3-3 and S3-4

In addition, three (3) side-by-side salvaged box beams decommissioned by Kent County Road

Commission after they have been in service for 39 years were tested in the Structural Testing

Laboratory at Lawrence Technological University for both nondestructive evaluation as well as

residual flexural failure testing. Prior to flexural testing, the following nondestructive methods

were deployed: ultrasonic assessment for delamination and void detection; electro-chemical half-

28  

cell assessment for detecting corrosive environment; impact hammer assessment of surfaces to

detect variations and potential delamination; and magnetic flux leakage to determine loss of cross

sectional area of reinforcement and strands. Details of all experimental procedures are presented

in subsequent sections.

Field inspections were also carried out on sections of the following existing bridges in Michigan,

US-131 over Muskegon River, southbound 6 miles south of Big Rapids and segments of I-96 over

Canal Road in Lansing. The employed nondestructive evaluation methods were ultrasonic,

magnetic flux leakage and electro-chemical half-cell assessments.

3.2. Construction of Two Samples of Specimen 1 Box Beams, S1-1 and S1-2

Two 15 feet long box beams were constructed with a cross section of 27 in x 36 in and in concept

designed according to MDOT Bridge Design Guides 6.65.10A with single layer of bottom

reinforcement. These laboratory beams represents box beam members in a typical reference bridge

with a span length of 60 feet and a deck width of 45 feet. The bridge would allow one lane of

traffic in each direction and two shoulders of 10 feet width. The laboratory beams were reduced

in length from 60 feet to 15 feet while still including a center transverse diaphragm and two cells.

This reduction enabled transporting the beams with the available lifts. Furthermore, due to the

limited scope of this project, no deck slab was incorporated during the construction of the beams,

and therefore, shear stirrups were not extended to integrate a slab system.

3.2.1 Design and Construction of Formwork.

The formwork was constructed mainly from plywood and consisted of base chairs, base-plates and

side plates. The base plate was supported on base chairs at a spacing of two feet along the length

of the base plate. Horizontal and vertical braces were also provided to support the side-plates and

to ensure straight alignment of the edges of the box-beams.

Interior cells were constructed using plywood boxes. The cells were fixed in place by suspending

them on threaded rods with end nuts outside the formwork. In the laboratory setting the final

placement of the cells were well controlled. For instance, one of the cells was intentionally shifted

horizontally towards one side of one box beam to create a thinner web. The intention was to

29  

evaluate the effectiveness of the phased array 3D ultrasonic tomography method to detect and

quantify defects in the thinner web. Formwork was designed and constructed in the laboratory by

research assistants and laboratory technical staff as shown in Figures 3.4 through 3.6.

Figure 3.4: Construction of formwork. Figure 3.5: Construction of interior cells

Figure 3.6: Formwork set-up

30  

3.2.2 Reinforcement Cages

All reinforcements were cut, bent and installed in the laboratory. Each box-beam cage was made

up of mild steel #4 steel stirrups at different spacing and five #4 non-prestressing bars provided

as top and bottom reinforcement according to MDOT Bridge Design Guides 6.65.10A. The center

of top reinforcements were located at 2.5 inches from the top surface of the beam. The mild steel,

longitudinal bars and stirrups, for the box beams reinforcement were Grade 60.

In addition, there were a total of 12 prestressing steel strands located at the bottom of the beam.

The steel prestressing strands were Grade 270 relaxed strands with diameter of 0.6 inch. The center

of the prestressing strands were located at 2 inches from the bottom surface of the beam. A total

prestressing force of 280 kips was symmetrically distributed among the bottom steel strands along

the cross section of each beam as follows: 4 strands were stressed 10 kips each, another 4 were

stressed 23 kips each, 2 out of the remaining 4 were stressed 30 kips each and the remaining 2

were stressed 44 kips. Details of the cross-section and longitudinal sections of the beams have

been provided in Figures 3.7 and 3.8. Figures 3.9 through 3.12 depicts the preparation and

installation of rebar for S1-1 and S1-2.

Figure 3.7: Cross-sectional details of box-beams

31  

Figure 3.8: Longitudinal details of box-beams

Figure 3.9: Bending stirrups Figure 3.10: Assembling of reinforcement

Figure 3.11: Reinforcement cages for box-beam Figure 3.12: Installing edge & side Panels

32  

3.2.3 Simulation of Defects

Defects of voids were simulated along the length of reinforcement to depict voids created around

reinforcement as a result of the corrosion process, see Figures 3.13 through 3.16. Furthermore, two

additional types of defects were used; three (3) simulated honeycombs and heavy grease around

several of the prestressing strands. The goal of these activities were to determine if these defects

could be detected using the phased array ultrasonic 3D tomography. The defect types and their

locations were documented as shown in Figures 3.17 through 3.19.

Fig. 3.13: Plastic defects around reinforcement Fig. 3.14: Plastic defects around bottom strand

Fig. 3.15: Honeycomb at the top of box beam Fig. 3.16: Plastic tubes for debonding

33  

Figure 3.17: Top plan of box beam showing locations and defect length around Top reinforcement

Figure 3.18: Bottom Plan of Box Beam 1 showing locations of defects around strands

Figure 3.19: South Elevation of Box Beam 1 showing locations of defects around strands

34  

3.2.4 Prestressing Steel Strands

End plates for the prestressing bed were redesigned to suit the proposed beam size and to fit for

the strands spacing as shown in Figure 3.20

Figure 3.20: End plates for prestressing bed. Figure 3.21: Prestressing steel strands

When construction of the steel cages and formwork were completed, the steel cages were placed

inside the formork. Plastic chairs with effective height of 1.5 in were attached to the underside of

the cages to provide the bottom cover. In addition, 1.25 in chairs were also attached to either side

of the cages to create the side concrete cover. Bulkheads were mounted at the two extreme ends of

the two beams. Conventional steel chucks were used as anchorage systems for the steel strands

and attached at both the live and dead-ends of the steel strands. Figure 3.21 depicts the application

of prestressing forces to the strands.

35  

3.2.5 Concrete Placement and Preparation of Cylinders.

Six (6) cubic yards of concrete with compressive strength of 7000 psi ordered from MCCOIG

Materials was used for the casting of the two (2) box beams and preparation of 24 concrete

cylinders for compressive strength testing as shown in Figures 3.22 and 3.23. Two slump tests

were conducted prior to casting of the box-beams and immediately after in accordance with ASTM

C143/C143-05 to check the workability of the concrete. Two electrical vibrators were used for

consolidation as well as mechanical rods around the diaphragm locations and the end-block of the

beams where the steel reinforcements were closely spaced.

Figure 3.22: Casting of Box Beams, S1-1 and S1-2

Fig. 3.23: Slump Test / Cylinders Preparation

36  

3.2.6 Curing Box Beams and Releasing of Steel Strands

The beams were cured under wet burlaps layered with plastic sheets to limit moisture loss and

surface cracking. Water was applied to the top beam surfaces for 7 days. Steel strands were

released after 4 days of curing by cutting them from the bulkheads in order to transfer the

compressive stresses to the concrete after it had gained adequate strength.

Figure 3.24: Curing Beams Figure 3.25: Releasing of steel strands.

3.2.7 Compressive Strength of Box Beams

The average 28 day compressive strength of the cylinders was 7,694 psi. The development of

average compressive strength with age and the summative data are shown in Table 1. The

compressive strength was determined according to ASTM C39 - 05, Standard Test Method for

Compressive Strength of Cylindrical Concrete Specimens. Figures 3.26 and 3.27 shows

compressive strength testing for 4 days and 7 days respectively, the prestressing force was

transferred at the 4 days compressive strength.

37  

Table 3.1: Compressive Strength Development using 6 inch by 12 inch Cylinders

Identification Max. Load Compressive Age of specimen Number (Pound-force) Strength (psi) (days)

C1 106,990 3,784 4

C2 118,320 4,185 4

C3 112,070 3,964 4

C4 142,470 5,039 7

C5 139,050 4,918 7

C6 147,400 5,213 7

C7 182,260 6,446 14

C8 201,780 7,137 14

C9 179,650 6,354 14

C10 207,050 7,323 28

C11 218,790 7,738 28

C12 226,750 8,020 28

Fig. 3.26: Four (4) Days at Transfer Fig. 3.27: 7 Days Compressive Strength

38  

3.2.8 Ultrasonic Assessment of S1-1 and S1-2.

Testing was done by scanning the entire geometry of the box beams to ascertain how correctly the

ultrasonic assessment predicts the thicknesses of the various components of the box beams. This

was used as a reference during scanning of strands, for instance to check on the location of strand

from the concrete surface a reference was made to the thickness of the bottom flange.

The following testing protocol was used for S1-1 and S1-2.

(i) Complete scan of all four longitudinal surfaces of each beam with varied frequencies of

30, 40, 50, 60, 70, 80, 90 and 100 kHz to examine how the different thicknesses of the

various components of the beam were predicted with the above frequencies.

(ii) Localized scan on defects: plastic tubes (denoting voids and debonding) and grease with

varied frequencies of 30, 40, 50, 60, 70, 80, 90 & 100 kHz to simulate assessment of voids

which usually occur around corroded strands and reinforcement as a result of corrosion.

(iii) Localized scan on defects: honey combs with frequencies of 30, 40, 50, 70, 80, 90 & 100

kHz to simulate deterioration in concrete structures such as cracks and delamination which

usually occurs as a result of corrosion of strands and reinforcement.

(iv) Localized scan on reinforcement and strands without defect, with frequencies of 30, 40,

50, 60, 70, 80, 90 & 100kHz were compared with localized scan on plastic defects to check

on the differences of the various scans.

(v) Localized scan on solid concrete sections without reinforcement, strand or any defect, with

frequencies of 30, 40, 50, 60, 70, 80, 90 & 100 kHz

Figures 3.28 through 3.30 show ultrasonic assessment from various locations along box beams

S1-1 and S1-2. Detailed analysis of the test results have been presented in sections 4.1 and 4.2.

 

 

 

   

39  

Figure 3.28: Scanning on the side of S1-1

Figure 3.29: Scanning on top of S1-2

Figure 3.30: Scanning on the bottom of S1-2

40  

3.3 Construction and Testing of Three Specimen 2 Beams

3.3.1 Construction of Specimen 2 Beams

3.3.1.1 Formwork and Preparation of Reinforcement/Strands

Formwork was designed and constructed in house in the laboratory by research assistants and

laboratory technical staff. Three sets of formwork were constructed to simulate three different

samples of Specimen 2 as shown in Figure 3.31. Reinforcement were cut and grinded to the

required percentage losses of cross-sectional areas as shown in Figure 3.32. The percentage losses

considered were 5, 10, 15, 20 and 30 with varying lengths of grinding along the reinforcement.

Two lengths of grinding configurations considered were 1inch-1inch-1inch and 5inches-2inches-

5inches as shown in Figures 3.33 and 3.34. Wires along 0.6 diameter strands were also cut varying

from one wire cut to four wires cut with varying cutting lengths of 0.25 inch to 1 inch and were

covered with dual-wall polyolefin plus duct tape as shown in Figure 3.36.

Fig. 3.31: Constructing formwork Fig. 3.32: Cross-sectional area loss on reinforcement

Fig. 3.33: Grinding configuration of 1(0.5+0.5)inch-1inch-1(0.5+0.5)inch for different

percentage losses for reinforcementbar

41  

Figure 3.34: Grinding configuration of 5(1+1+1+1+1) inches-2 inches-5(1+1+1+1+1) inches

for different percentage losses for reinforcement

Figure 3.35: 0.6” Strand length of wire cut Figure 3.36: Wire cut covered with dual wall

Polyolefin tube and duct tape

3.3.1.2 Preparing the cages

Reinforcemnet and strands were tied with documented defect locations as shown in

Figures 3.37 through 3.39. Three sets of cages were formed, one for single layer of top and bottom

reinforcement with defects in both top and bottom reinforcement, a second with double bottom

layer of reinforcement with defects in both top reinforcement as well as second layer of

reinforcemnt from the bottom. A third was formed for single layer of top strands with double layer

of bottom strands with defects in the top strands and the second layer of strands from the bottom.

The reinforcement grinded to simulate cross-sectional area losses were #5 steel of Grade 60 with

#3 stirrups of Grade 60. The longitudinal prestressing strands with wire cuts to simulate cross-

sectional area losses were 0.6 in, low-relaxation seven-wire steel strands of Grade 270.

Length of cut in steel strand

Plastic tube & duct tape

42  

Figure 3.37: Cages of reinforcement and strands Figure 3.38: Cross-sectional Area loss

Figure 3.39: Strands showing loss of wires covered with

Dual wall polyolefin tube and duct tape

3.3.1.3 Placing of Concrete and Preparation of Concrete Cylinders

Two cubic yards of concrete with compressive strength of 7000 psi ordered from MCCOIG

Materials was used for the casting of three (3) Specimen 2 beams and preparation of cylinders for

compressive strength testing as shown in Figures 3.40-43.

Wire cut defect 

Loss in cross sectional area  

43  

Fig. 3.40: Placing of concrete Fig. 3.41: Leveling top of specimen 2 beams

Figure 3.42: Checking for slump Figure 3.43: Preparation of concrete cylinders

3.3.2 Ultrasonic Assessment of S2-1, S2-2 and S2-3.

Ultrasonic testing of Specimen 2 beams was undertaken using the following testing protocol.

(i) Complete scan of top and bottom surfaces of each of S1-1, S2-2 and S2-3 with varied

frequencies of 30, 40, 50, 60, 70, 80, 90 and 100 kHz to examine how the different

cross-sectional area losses due to grinding of reinforcement and cutting of wires of strands

were detected with the above frequencies.

(ii) Localized scans on defect locations with varied frequencies of 30, 40, 50, 60, 70, 80, 90 & 100

kHz

44  

(iii) Localized scan on reinforcement and strands without defect, with frequencies of 30, 40, 50,

60, 70, 80, 90 & 100 kHz.

Details of reinforcement configurations for S2-1, S2-2 and S2-3 with documented locations of

these simulated defects are shown in Figures 3.44 through 3.52. Figure 3.53 shows ultrasonic

assessment set-up for S2-1. Figures 3.54 and 3.55 show typical scan on reinforcement with defect

as well as scan on reinforcement without defect.

45  

Figure 3.44: Top Reinforcement Arrangement and Percentage Losses for S2-1 with Total Length of 3 inches

Cross-sectional Area Loss

46  

Figure 3.45: Bottom Reinforcement Arrangement and Percentage Losses for S2-1 with Total length of 3 inches

Cross-sectional Area Loss

47  

Figure 3.46: Cross-sectional Area of S2-1 showing Single Bottom Reinforcement

48  

Figure 3.47: Top Reinforcement Arrangement and Percentage Losses for S2-2 with Total Length of 12 inches

Cross-sectional Area Loss

49  

Figure 3.48: Bottom Reinforcement Arrangement and Percentage Losses for S2-2 with Total Length of 12 inches

Cross-sectional Area Loss

50  

Figure 3.49: Cross-sectional Area of S2-2 showing Double Bottom Reinforcement

51  

Figure 3.50: Top Strand Arrangement, Length and Number of Wire cuts for S2-3

52  

Figure 3.51: Bottom Strand Arrangement, Length and Number of Wire cuts for S2-3

53  

Figure 3.52: Cross-sectional Area of S2-3 showing Double Bottom Strands

54  

Figure 3.53: Ultrasonic assessment set-up for S2-1.

Figure 3.54: Scan on Reinforcement with defect at frequency of 70 kHz with dB of 33.8

Reflection from reinforcement with

cross-sectional area loss

Back wall reflection from underside of beam

55  

Figure 3.55: Scan on reinforcement without defect at frequency of 70 kHz with dB of 33.8

3.3.3 Calibration of Magnetic Flux Leakage (MFL) Mobile Scanner

Calibration of the MFL Mobile scanner was done, using undamaged reinforcement with 10%, 20%

and 30% loss of cross-sectional area for reinforcement #4 - #9. Two timber formworks of size

48 inches x18 inches x 11inches were constructed with three (3) 2 inches holes punched on two

opposite sides of the shorter faces to calibrate the influence of the magnetic system on the

reinforcement at increasing depths away from the magnetic influence as shown in Figure 3.56. The

first formwork had holes at increments of 2.5 inches, while the second had holes at depths of

1.5 inches, 3 inches, and 6 inches. A baseline reading of 2.935 volts from the middle Hall-effect

sensor, M3, was taken by running the scanner over the formwork with no reinforcement present.

Undamaged reinforcement #4 through #9 each were scanned ten (10) times at each depth

increments. The voltage reading from the middle hall-effect sensor, M3, was averaged for each

run and then all ten runs were averaged together.

The process was repeated for reinforcement #4 - #9 with the center foot of the reinforcement

gradually milled to a 10%, 20% and 30% loss in cross-section area. The max voltage reading,

which indicates the peak voltage in the affected area, from the middle hall-effect sensor, M3, was

averaged for each run and then all ten runs were averaged together.

Reflection from reinforcement without

cross-sectional loss

Back wall reflection from underside beam

56  

Figure 3.57 shows the trend line of the baseline adjusted average peak voltage for each depth. The

graph in Figure 3.57 shows the trend line for each depth up to three inches. Figure 3.58 shows a

scan of reinforcement #7 at the 1.5 inch depth. There is an increase in voltage as the MFL mobile

scanner passes over the affected area. This is due to more magnetic flux reaching the Hall-effect

sensor array than would be the case if the reinforcement did not have cross-sectional loss. Figure

3.59 shows magnetic flux voltage change of #7 reinforcement at 1.5 inches depth with cross-

sectional area loss of 20%.

   

 

                                                                                                    

Figure 3.56: MFL Mobile Scanner Calibration Setup

57  

Figure 3.57: Magnetic flux leakage influence with depth.

y = 116.36x0.2332

R² = 0.91y = 110.88x0.3616

R² = 0.96

y = 62.981x0.2571

R² = 0.69

y = 62.545x0.1826

R² = 0.45

y = 42.294x0.1111

R² = 0.50

0

20

40

60

80

100

120

140

0.00 0.20 0.40 0.60 0.80 1.00 1.20

Hall V

oltage (‐m

V)

Reinforcement Cross Section Area (in squared)

Effect of Reinforcement Depth Below Surface on Hall Voltage Change for Increasing Rebar Size

Depth2.5inches

Depth 3inches

Depth 5inches

Depth 6inches

Depth7.5inches

58  

Figure 3.58: Graph of voltage change for each depth

 

Figure 3.59: Graph showing effect of 20% loss

2.45

2.50

2.55

2.60

2.65

2.70

2.75

2.80

2.85

2.90

2.95

0 5 10 15 20

Hall V

oltage (V)

Distance along Surface with MFL Centered Directly Above Rebar (Inches)

#7 Reinforcement 1.5 inch Depth from Surface, 20 % Cross Section Loss

Change in Vertical Magnetic Flux Leakage

Change in Horizontal Magnetic Flux Leakage

M3 baseline reading: 2.935 Volts

‐60

‐40

‐20

0

20

40

60

80

0 5 10 15 20

Hall V

oltage Chan

ge (mV)

Distance along Surface with MFL Centered Directly Above Rebar (Inches)

Magnetic Flux Voltage Change#7 Reinforcement, 1.5 inch Depth, 20% Cross Section Loss

Vertical (M3)

Horizontal (F8)

Change in mV proportional to change in cross section area